What is your newsroom already doing that your audience might want to hear on Alexa?
You might already be producing something your audience wants, but you should let them tell you what it is
Since its 2014 launch, Amazon’s Echo smart speaker has turned its virtual audio assistant, Alexa, into a new member of the family. If you’re wondering about how many families Alexa has joined, Amazon said in October 2017 that more than 20 million Echo products had been sold, and a report issued in January indicates that Alexa is getting more popular.
In an attempt to be invited into these families’ homes, news organizations are developing what Amazon calls “skills” for Alexa to be part of her Flash Briefing. As NPR, CNN, Washington Post, AP and others share updates on Alexa, your newsroom might be wondering if it can get in the game too. My experience creating an Alexa skill for Annenberg Media, a student-led news organization at the University of Southern California, taught me that your newsroom might already have something that works, but you should let your audience tell you what it is. Here are the takeaways from our process.
1. Listen to your audience’s needs, not your own
During the first few meetings with the Alexa team, we floated several ideas: a Q&A format, a USC briefing aimed at parents and a quick rundown of Annenberg Media’s top headlines. We were discussing ideas that felt achievable with our resources and workflow and that were based on assumptions, such as that more parents than students might have Echo devices. Of course, we quickly realized that the briefing wasn’t for us — it was for our audience. So, we conducted a survey to figure out what they wanted to hear from Annenberg Media through an Alexa skill.
I created a Google Form asking current Alexa users in our community about how they use their Echo device. I asked things such as, “Which news briefings do you listen to?”, “What is your ideal amount of time for a briefing from one news organization?” and “What would make you want to listen to USC news on Alexa?” The goal was to figure out who our potential Alexa audience was and what they want. To get responses, the team distributed the survey through Annenberg Media’s social platforms.
We received 23 responses within a week, most of them from students at USC — an important point about who we had the potential to reach with an Alexa skill. And while it wasn’t much data, we at least had some idea of what our audience wanted.
What we gathered from the survey was that students were using Echo products, they wanted a round-up of USC news, events on campus and unique USC stories, and they wanted their briefing to be five minutes or less.
2. Outside feedback might prompt you to look within
Even though we were thinking about our process and resources during our initial brainstorming, it took feedback from our audience to point us to something the newsroom was already producing: a weekly newsletter and Snapchat show called The Rundown. Those products include, you guessed it, a roundup of USC news and events on campus. They don’t exactly have unique USC stories, though The Rundown newsletter ends with an “LOL” section that is usually a USC-focused meme.
So we ended up with an Alexa skill that is another version of The Rundown. Instead of the visual “LOL” that ends the newsletter, the Alexa skill satisfies the request from our audience for unique USC stories by ending with an existing audio series called Trojan Tales, which focuses on little-known stories about USC. We have different hosts than the Snapchat show, and obviously we don’t focus on stories with strong visuals like the show does. But the concept remains the same: a USC-focused weekly news roundup. Here’s an example The Rundown for Alexa:
An important caveat: Good products don’t come from looking at your own workflow. They come from focusing on the needs of others. That our audience pointed us to The Rundown was a happy accident. I’m not recommending you simply adapt something from within your newsroom. I’m recommending that you analyze your audience’s needs, then decide that if applies to an existing newsroom product.
3. Your adaptation needs to consider Alexa’s native features
We decided that the Alexa version of The Rundown would be published once a week, on Fridays, without considering one important detail: Since you prompt a Flash Briefing as a part of a package on Alexa, there wasn’t a good way for people to seek out The Rundown on Fridays only. So it will pop up every day of the week if we don’t replace it with a new episode.
The solution: We created an away message for the majority of the week (Monday through Thursday) in place of the briefing so that users wouldn’t be listening to the same episode if they were calling on the skill more than once a week. (Another hiccup for us is that our newsroom isn’t in production over the summer, so don’t try to listen to The Rundown now. We created an away message for summer, but after about a week, Alexa tells the audience that there’s no recent content. We’re working on it, but this shouldn’t be a problem for your newsroom.)
Before you run off and start a skill, remember: Storytelling on emerging platforms should be intentional
We officially launched our skill on April 13, three months after the project was pitched. So this project wasn’t something we decided to do on a whim. It took a good chunk of the semester to develop content for the skill and figure out what worked best for our audience. We took the time to learn if they were using Echo products, what they were already listening to, and how we could cater to what they wanted to hear from Annenberg Media. Most importantly, we created something that was useful to the user, and we followed up our experiment by looking at the metrics. Here’s a snapshot of what we found:
In the last 30 days, we’ve had 22 unique listeners and a total of 94 plays. At the time this data was recorded, we distributed three briefings and the two away messages through Alexa after the launch. I didn’t have time to do more data or consumer testing, but this was a great start to see where our audience is and whether or not they were listening.
Each unique listener has listened to the briefings at least four times, which actually aligns with the data from our survey. The number of unique listeners isn’t the minuscule number that I thought it would be, which gives some support to the analysis of our survey results: USC students want a news briefing tailored to them.
While testing news distribution on emerging platforms isn’t unusual for Annenberg Media, it shouldn’t be a deterrent for any other student-led or professional news organization. This project taught us that a good product doesn’t have to be restricted to just a few desks in the newsroom, it can be as versatile as you want it to be — as long as you listen to your audience.